CN103870659A - Failure analysis method for numerically-controlled machine tool - Google Patents

Abstract

The invention discloses a failure analysis method for a numerically-controlled machine tool. The failure analysis method overcomes the defects that in the prior art, failure relevance is not considered in machine tool failure analysis. A DEMATEL-ISM method is integrated, related failure statistics data are combined, failure correlation between subsystems is taken into consideration, a digraph and matrix operations are applied to obtain a comprehensive influence matrix between the subsystems and relevancy, an overall influence matrix and a reachable matrix are obtained through the comprehensive influence matrix between the subsystems, and the reachable matrix is decomposed so that a multilevel hierarchical structure model can be obtained. The relevancy and the multilevel hierarchical structure model are synthesized to obtain a numerically-controlled machine tool key subsystem; possible failure modes of components of the key subsystem and influences of the failure modes on operation of a numerically-controlled machine tool are determined by means of FMECA technical analysis, a single point of failure is found, and perniciousness of the failure modes is determined according to the severe degree of the failure modes and the probability of occurrence of the failure modes.

Description

A kind of fault of numerical control machine tool analytical approach

Technical field

The invention belongs to numerical control machine tool technique field, relate to a kind of fault of numerical control machine tool analytical approach, be specifically related to a kind of fault of numerical control machine tool analytical approach of considering numerically-controlled machine subsystem fault correlativity.

Background technology

Along with the enhancing of system compositeness function and advanced raising, the structure of system is also increasingly sophisticated, and each subsystem fault potential risk of numerically-controlled machine increases.In the normal use procedure of numerically-controlled machine, each numerically-controlled machine subsystem breaks down and all can cause numerically-controlled machine machine failure, and due to the existence of fault correlationship, unit or subsystem break down, may cause other parts of system to break down, cause Domino effect, form out-of-service sequence and inefficacy snowslide.Therefore, consider that fault correlationship carries out system Reliability Research and become more and more urgent.And this point has been ignored in the fault analysis of existing numerically-controlled machine research conventionally.

Exist in the system of dependent failure several unit interdependent each other, brought very big difficulty to systems reliability analysis.To fault, relevant research mainly concentrated in the research of unidirectional dependent failure in recent years, had analysis of series connection dependent failure analysis based on reliability model, negative correlation fault analysis, common cause failure etc.Research for bi-directional failure correlativity is scarcely out of swaddling-clothes, and is mainly the analysis for the degree of correlation between key element.Have document to consider that from the number of stoppages angle parts are to complete machine influence degree, but the method belong to single factor analysis method, and correlationship and the impact of each trouble location and fault mode between same level or different levels are difficult to comprehensive assessment.Have document utilization copula function to solve correlation coefficient value, but the method can not specify interaction relationship and action direction between subsystem.There is document to adopt autocorrelation matrix to consider that between each key element, direct relation has but been ignored the indirect correlation relation between key element in multi-level fault delivery chain.

Summary of the invention

Technical matters to be solved by this invention has overcome the problem of not considering failure dependency in the machine failure analysis that prior art exists, and a kind of fault of numerical control machine tool analytical approach of considering fault of numerical control machine tool correlativity is provided.

For solving the problems of the technologies described above, the present invention adopts following technical scheme to realize, and accompanying drawings is as follows:

The present invention is the fault of numerical control machine tool analytical approach of numerically-controlled machine in the time considering subsystem fault correlativity.Can realize the fault of numerical control machine tool correlation analysis under relevant of fault subsystem fault in numerically-controlled machine tool system operational process, for judging key subsystem supplying method, for failure dependency modeling and Reliability Distribution lay the foundation, and provide new method for Fault Quick Diagnosis and location, simultaneously by vertical analysis, realize obtaining from key subsystem to critical failure pattern, the clear and definite concrete improvement direction of reliability.

This fault of numerical control machine tool analytical approach, in conjunction with fault statistics related data, consider fault subsystem correlationship, adopt decision experiments chamber analytic approach (Decision Making Trial and Evaluation laboratory, DEMATEL), application digraph and matrix operation obtain fault subsystem combined influence relational matrix and the degree of correlation; And integrated ISM (Interpretative Structural Modeling, ISM), interactional numerous key elements logical relation is carried out to visual representation by Multilevel Hierarchical hierarchy Model, and synthesis pertinence and Multilevel Hierarchical hierarchy Model obtain numerically-controlled machine key subsystem.For further determining the critical failure pattern of key subsystem, utilize FMECA analytical technology to carry out HAZAN to key subsystem, thereby find out critical failure pattern, specify reliability improvement emphasis.The system applies of the method is all significant to reliability considerations such as fault diagnosis and location, system reliability improvement.

About DEMATEL-ISM method:

DEMATEL method is that American national laboratory in 1971 is to solve complicated in real world, difficult problem proposes a kind of methodology of carrying out factor of system analysis by graph theory and matrix tool.It is by the logical relation between each key element in analytic system, build direct influence matrix, through calculation, obtain degree of affecting, the degree of being affected of each key element on other key elements, and reason degree and the centrad of each factor, and then obtain affecting between key element relation and degree.

ISM method was set it as for a kind of analytical approach of complex socioeconomic system relevant issues and is proposed by U.S. professor Warfield in 1973.The method ultimate principle is to adopt various creative technology, carry out the extraction of problem inscape, by instruments such as digraph, matrix or computing machines, System Construction becomes a Multilevel Hierarchical hierarchy Model the most at last, dependence between key element and internal system structure are shown intuitively to the methodization of implementation relation, stratification.

For simplifying relation between clear and definite factor of system and simplifying the matrix operation of large amount of complex in ISM modeling, the present invention will quote integrated DEMATEL-ISM method numerically-controlled machine will be carried out to the relevant relationship analysis of fault.First, set up fault subsystem digraph according to subsystem fault related data; Secondly, calculate the combined influence matrix T between subsystem by DEMATEL method, draw accordingly degree of impact, degree of being affected, centrad and the reason degree of the each subsystem of numerically-controlled machine, thereby obtain fault subsystem relevancy ranking; Then, determine entire system influence matrix H, H=T+I, and obtain reachability matrix M according to the mapping relations between entire effect matrix H and reachability matrix M; Finally, in conjunction with ISM method, system is carried out to grade classification and Region Decomposition, obtain the hierarchical structure relation between fault subsystem.Two kinds of methods are integrated, both can specify subsystem fault impact size, also can clear failure transport mechanism.

A kind of fault of numerical control machine tool analytical approach, comprises the following steps:

Step 1: integrated DEMATEL-ISM method analysis of failure subsystem influences each other, and obtains the fault subsystem degree of correlation and Multilevel Hierarchical hierarchy Model, and then obtains key subsystem;

In conjunction with fault statistics related data, consider fault correlationship between subsystem, adopt Decision Making Trial and Evaluation laboratory decision experiments chamber analytic approach, application digraph and matrix operation obtain fault subsystem combined influence matrix and the degree of correlation, and apply Interpretative Structural Modeling ISM, the interactional logical relation of numerous fault subsystems is carried out to visual representation by Multilevel Hierarchical hierarchy Model, the resultant fault subsystem degree of correlation and Multilevel Hierarchical hierarchy Model obtain numerically-controlled machine key subsystem,

Step 2: utilize FMECA analytical technology to carry out fault analysis to gained key subsystem and try to achieve critical failure pattern;

The fault mode that may exist by the each ingredient of Analysis deterrmination numerically-controlled machine key subsystem and the impact of each fault mode on numerically-controlled machine, find Single Point of Faliure, and according to each fault mode severity and each fault mode probability of happening, determine each fault mode harmfulness, for taking correspondingly to remedy innovative approach to eliminate or to alleviate each fault mode harmfulness foundation is provided.

Integrated DEMATEL-ISM method analysis of failure subsystem described in technical scheme influences each other, and obtains the fault subsystem degree of correlation and Multilevel Hierarchical hierarchy Model, and then obtains key subsystem, comprises the following steps:

Step 1: according to fault statistics related data, build fault digraph, determine that the fault phase that forms numerically-controlled machine closes set of subsystems S={s _i, i=1,2 ..., n; Wherein S _irepresent that i and other subsystems exist the subsystem of dependent failure, n represents to exist the fault subsystem number of dependent failure;

Step 2: fault digraph is converted into the direct influence matrix Y between all fault subsystems;

Step 3: to the direct influence matrix Y standardization between all fault subsystems;

Step 4: calculate combined influence matrix T

$T=\underset{k\&RightArrow;\&infin;}{\lim }(X+{\mathrm{<figure-callout\; id="2"\; label="X"\; filenames="HDA0000483640660000011.png,HDA0000483640660000021.png"\; state="\{\{state\}\}">X</figure-callout>}}^2+\&CenterDot;\&CenterDot;\&CenterDot;X^k)=X{(I-X)}^{-1}$

In formula: I is unit matrix, X ^krepresent the k stage remote effect of subsystem i to subsystem j, X is standardization matrix;

Step 5: degree of impact, degree of being affected, centrad, the reason degree of obtaining each fault correlation subsystem;

Step 6: determine reachability matrix M;

Combined influence matrix T only reflects the relation that influences each other and the degree between different faults subsystem, does not consider the impact of fault subsystem on self, therefore needs to calculate the entire effect relation of faults subsystem;

Step 7: Region Decomposition and level are divided: definition is subject to subsystem S _ithe set of subsystems of impact is subsystem S _ireachable set R (S _i), affect subsystem S _iset of subsystems be subsystem S _ithe A of collection in advance (S _i), by reachable set R (S _i) and collect in advance A (S _i) calculating, reachability matrix M is carried out to Region Decomposition and level and divides;

Step 8: the reachability matrix removal of binary relation of bypassing the immediate leadership to step 7 after level is divided, and remove the binary relation self arriving, obtain skeleton matrix, and then obtain Multilevel Hierarchical hierarchy Model, thereby obtain key subsystem.

Direct influence matrix Y fault digraph being converted between all fault subsystems described in technical scheme refers to: failure dependency is taken into account to Y=(y _ij) _{n × n}; Wherein y _ijfor subsystem i affects the number of times that affects of subsystem j, when i=j, y _ii=0, n represents to exist the fault subsystem number of dependent failure;

The direct influence matrix Y standardization between all fault subsystems described in technical scheme refers to: obtain standardization matrix X

$X=\frac{Y}{\max \left(\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{y}_{\mathrm{ij}}\right)}$

In formula: y _ijfor subsystem i affects the number of times that affects of subsystem j, n represents to exist the fault subsystem number of dependent failure;

The degree of impact of obtaining each fault correlation subsystem described in technical scheme, degree of being affected, centrad, reason degree refer to: row and the vector of establishing combined influence matrix T are Q, row and the vectorial D of being,

T=[t _ij] _n×n,i,j=1,2,…n

$Q_i=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{t}_{\mathrm{ij}},i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;n$

$D_i=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{t}_{\mathrm{ij}},j=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;n$

In formula: Q _irepresent degree of impact; D _irepresent degree of being affected; t _ijrepresent the direct and remote effect degree size of subsystem i to subsystem j, t _ij≠ 0, illustrate that subsystem i and subsystem j have the fault relation that influences each other, otherwise irrelevant; N represents to exist the fault subsystem number of dependent failure;

Therefore, Q _i+ D _jcentered by degree, Q _i-D _jfor reason degree:

$Q_i+D_i=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{t}_{\mathrm{ij}}+\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{t}_{\mathrm{ij}},i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;n$

$Q_i-D_i=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{t}_{\mathrm{ij}}+\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{t}_{\mathrm{ij}},i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;n;$

The entire effect relation of calculating faults subsystem described in technical scheme refers to: calculate entire effect matrix H, its computing formula is:

H=T+I=[h _ij] _n×n

In formula: I is unit matrix, h _ijafter representing to consider that fault subsystem affects self, the direct and remote effect degree size of subsystem i to subsystem j, n represents to exist the fault subsystem number of dependent failure;

By entire effect matrix H, can determine reachability matrix M, order

M=[m _ij] _n×n,i,j=1,2,…n

M _ijvalue according to the following formula:

wherein λ is given threshold value, and the size of λ directly affects reachability matrix formation and follow-up hierarchical structure is divided; For the less system of n value, without simplification, λ=0 is set;

M _ijbe illustrated under given threshold value λ, whether subsystem i there is impact to subsystem j, if h _ij> λ, represents to exist impact, m _ijvalue is 1; If h _ij≤ λ, represents not exist impact, m _ijbe 0;

Region Decomposition described in technical scheme refers to: by R (S _i) ∩ A (S _i) be defined as common collection, common collection R (S _i) ∩ A (S _i)=A (S _i) key element be defined as initial set B (S), for the S in B (S) _iand S _jif, R (S _i) ∩ A (S _j)=Φ, S _iand S _jdo not belong to the same area, otherwise be the same area, decomposition that so can feasible region; The result of Region Decomposition can be designated as P (S)=P ₁, P ₂... P _k, wherein P _kit is the set of subsystems in k relatively independent region;

Level described in technical scheme is divided and is referred to: for the same area P ₁, obtain and meet R (S successively _i) ∩ A (S _i)=R (S _i) subsystem, find out each class set, be expressed as

L ₁={S _i/S _i∈P ₁-L ₀,R(S _i)∩A(S _i)=R(S _i)i,=1,…2,n}

_L2={S _i/S _i∈P ₁-L ₁,R(S _i)∩A(S _i)=R(S _i)i,<n}

…

L _k={S _i/S _i∈P ₁-L ₁-…-L _k-1,R(S _i)∩A(S _i)=R(S _i)i,<n}

In formula: L _kfor level is divided the each class set drawing;

Described in technical scheme, utilize FMECA Technique of Qualitative Analysis to carry out fault analysis to gained key subsystem to try to achieve critical failure pattern, comprise the following steps:

Step 1: Failure Mode Effective Analysis; Be related in situation at known key subsystem reliability logic, by fault statistics data, the fault mode that key subsystem parts are likely occurred is added up, and analyzes impact and consequence that each fault causes numerically-controlled machine;

Step 2: severity is determined; Final influence degree according to fault mode in fault data to numerically-controlled machine, determines the severity of each fault mode;

Step 3: fault mode probability of happening is determined; From the fault statistics data of key subsystem, analyze and obtain this subsystem fault pattern probability;

Step 4: fill in FMECA form; The combined influence that the each fault mode severity of key subsystem and fault mode probability of happening are produced is classified, and by filling in FMECA form, the pattern impact of likely breaking down is carried out to comprehensive comprehensive evaluation;

Step 5: harmfulness matrix analysis; Harmfulness matrix horizontal ordinate is fault mode severity grade, ordinate is the probability grade of fault mode, put the distance of initial point by the vertical projection of each fault mode distributed points on diagonal line and carry out comparison harmfulness size, this fault mode harmfulness of the longer explanation of distance is larger.

The severity of the definite each fault mode described in technical scheme refers to: severity is divided into 4 ranks, be respectively I critical fault, II catastrophic failure, III generic failure, IV minor failure, wherein I grade failure effect damages for causing numerically-controlled machine, II grade failure effect is to cause heavy economic losses or cause the numerically-controlled machine of mission failure badly damaged, III grade failure effect is to cause certain economic loss or cause the numerically-controlled machine of task delay or degradation slightly to damage, IV grade failure effect is for being not enough to cause above-mentioned 3 kinds of consequences, but can cause unscheduled maintenance or repairing.

Analysis described in technical scheme obtains this subsystem fault pattern and refers to: fault mode probability of happening grade is divided into A, B, C, D, an E5 grade, represents respectively often to occur, sometimes occur, occur once in a while, seldom occur, seldom occur; Fault mode probability of happening feature be followed successively by fault mode probability of happening account for more than 20% high probability of total failare probability, be greater than total failare probability 10% or be less than or equal to its middle equiprobability of 20%, be greater than 1% of total failare probability and be less than or equal to its 0.1% of total failare probability that seldom occurs, is greater than of 10% and be less than or equal to its unlikely generation of 1%, account for total failare probability 0.1% or following almost nil.

Compared with prior art the invention has the beneficial effects as follows:

1. the present invention is building on the basis of fault subsystem digraph, adopts DEMATEL method to carry out computational analysis to the degree of correlation of machining center dependent failure subsystem.The position can failure judgement being taken place frequently by centrad numerical value, judgement and the tight subsystem of other fault subsystems.Can determine the maximum source of trouble by reason degree, degree of impact and degree of being affected value, and can specify the dependent failure direction of propagation, determine driver sub-system and executive subsystem, for failure dependency modeling and Reliability Distribution lay the foundation.

2. the present invention is the deficiency that makes up independent use DEMATEL, and integrated ISM method has built fault correlation subsystem multilevel hierarchical structure model, makes fault delivery relation be able to performance directly perceived and deep, provides new method for carrying out Fault Quick Diagnosis and location.

3. the present invention adopts FMECA to carrying out criticality analysis based on DEMATEL-ISM gained key subsystem, has found out critical failure pattern, the clear and definite concrete improvement direction of reliability.

4. the present invention provides the new approaches of a set of simple and fast and system for system trouble analysis, perfect existing fault analysis theory and method system.

Accompanying drawing explanation

Below in conjunction with accompanying drawing, the present invention is further illustrated:

Fig. 1 is the numerically-controlled machine dependent failure subsystem fault digraph in a kind of fault of numerical control machine tool analytical approach of the present invention;

Fig. 2 is the centrad of the numerically-controlled machine dependent failure subsystem in a kind of fault of numerical control machine tool analytical approach of the present invention and the reason line chart of writing music;

Fig. 3 is the numerically-controlled machine Multilevel Hierarchical hierarchy Model figure in a kind of fault of numerical control machine tool analytical approach of the present invention;

In figure: S ₁for electrical system, S ₂for tool magazine, S ₃for feed system, S ₄for cooling system, S ₅for chip removal system, S ₆for lubricating system, S ₇for digital control system, S ₈for axis system, S ₉for hydraulic system, Q _i+ D _jcentered by degree, Q _i-D _jfor reason degree, S _ifor dependent failure subsystem, L ₁for top layer fault subsystem, L ₂for shallow-layer fault subsystem, L ₃for middle level fault subsystem, L ₄for deep layer fault subsystem.

Embodiment

Below in conjunction with accompanying drawing, the present invention is explained in detail:

A kind of fault of numerical control machine tool analytical approach, comprises the following steps:

(1) integrated DEMATEL-ISM method analysis of failure subsystem influences each other, and obtains the fault subsystem degree of correlation and Multilevel Hierarchical hierarchy Model, and then obtains key subsystem;

Step 1: according to fault statistics related data, build fault digraph, determine that the fault phase that forms numerically-controlled machine closes set of subsystems S={s _i, i=1,2 ..., n; Wherein S _irepresent that i and other subsystems exist the subsystem of dependent failure, n represents to exist the fault subsystem number of dependent failure; With certain series machining center fault statistics data instance, consult Fig. 1, can obtain concrete relevant factor;

Step 2: fault digraph is converted into the direct influence matrix Y between all fault subsystems; Failure dependency is taken into account to Y=(y _ij) _{n × n}; Wherein y _ijfor subsystem i affects the number of times that affects of subsystem j, when i=j, y _ii=0, n represents to exist the fault subsystem number of dependent failure;

Step 3: to the direct influence matrix Y standardization between all fault subsystems, obtain standardization matrix X

$X=\frac{Y}{\max \left(\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{y}_{\mathrm{ij}}\right)}$

In formula: y _ijfor subsystem i affects the number of times that affects of subsystem j, n represents to exist the fault subsystem number of dependent failure;

Step 4: calculate combined influence matrix T

$T=\underset{k\&RightArrow;\&infin;}{\lim }(X+{\mathrm{<figure-callout\; id="2"\; label="X"\; filenames="HDA0000483640660000011.png,HDA0000483640660000021.png"\; state="\{\{state\}\}">X</figure-callout>}}^2+\&CenterDot;\&CenterDot;\&CenterDot;X^k)=X{(I-X)}^{-1}$

In formula: I is unit matrix, X ^krepresent the k stage remote effect of subsystem i to subsystem j, X is standardization matrix;

Step 5: degree of impact, degree of being affected, centrad, the reason degree of obtaining each fault correlation subsystem;

If the row of combined influence matrix T and vector are Q, row and vector is D,

T=[t _ij] _n×n,i,j=1,2,…n

$Q_i=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{t}_{\mathrm{ij}},i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;n$

$D_i=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{t}_{\mathrm{ij}},j=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;n$

In formula: Q _irepresent degree of impact; D _irepresent degree of being affected; t _ijrepresent the direct and remote effect degree size of subsystem i to subsystem j, t _ij≠ 0, illustrate that subsystem i and subsystem j have the fault relation that influences each other, otherwise irrelevant; N represents to exist the fault subsystem number of dependent failure;

Therefore, Q _i+ D _jcentered by degree, Q _i-D _jfor reason degree:

$Q_i+D_i=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{t}_{\mathrm{ij}}+\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{t}_{\mathrm{ij}},i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;n$

$Q_i-D_i=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{t}_{\mathrm{ij}}+\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{t}_{\mathrm{ij}},i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;n;$

Because t _ijdirect and the remote effect degree size of the subsystem i referring to subsystem j, i.e. combined influence degree size, if t _ij≠ 0, illustrate that subsystem i and subsystem j have the fault relation that influences each other, t _ij=0 illustrates both

Combined influence degree is 0, does not have the fault relation that influences each other, and that is to say and does not have fault correlationship;

For making above numerical value performance more directly perceived, draw failure branch chart, consult Fig. 2.

Step 6: determine reachability matrix M;

Combined influence matrix T only reflects the relation that influences each other and the degree between different faults subsystem, do not consider the impact of fault subsystem on self, therefore need to calculate the entire effect relation of faults subsystem, i.e. entire effect matrix H, its computing formula is:

H=T+I=[h _ij] _n×n

In formula: I is unit matrix, h _ijafter representing to consider that fault subsystem affects self, the direct and remote effect degree size of subsystem i to subsystem j, n represents to exist the fault subsystem number of dependent failure;

By entire effect matrix H, can determine reachability matrix M, make M=[m _ij] _{n × n}, i, j=1,2 ... n

M _ijvalue according to the following formula:

wherein λ is given threshold value, and the size of λ directly affects reachability matrix formation and follow-up hierarchical structure is divided; For the less system of n value, without simplification, λ=0 is set;

M _ijbe illustrated under given threshold value λ, whether subsystem i there is impact to subsystem j, if h _ij> λ, represents to exist impact, m _ijvalue is 1; If h _ij≤ λ, represents not exist impact, m _ijbe 0;

Step 7: Region Decomposition and level are divided: definition is subject to subsystem S _ithe set of subsystems of impact is subsystem S _ireachable set R (S _i), affect subsystem S _iset of subsystems be subsystem S _ithe A of collection in advance (S _i), by reachable set R (S _i) and collect in advance A (S _i) calculating, reachability matrix M is carried out to Region Decomposition and level and divides;

Region Decomposition: by R (S _i) ∩ A (S _i) be defined as common collection, common collection R (S _i) ∩ A (S _i)=A (S _i) key element be defined as initial set B (S), for the S in B (S) _iand S _jif, R (S _i) ∩ A (S _j)=Φ, S _iand S _jdo not belong to the same area, otherwise be the same area, decomposition that so can feasible region; The result of Region Decomposition can be designated as P (S)=P ₁, P ₂... P _k, wherein P _kit is the set of subsystems in k relatively independent region;

Level is divided: for the same area P ₁, obtain and meet R (S successively _i) ∩ A (S _i)=R (S _i) subsystem, find out each class set, be expressed as

L ₁={S _i/S _i∈P ₁-L ₀,R(S _i)∩A(S _i)=R(S _i)i,=1,…2,n}

L ₂={S _i/S _i∈P ₁-L ₁,R(S _i)∩A(S _i)=R(S _i)i,<n}

…

L _k={S _i/S _i∈P ₁-L ₁-…-L _k-1,R(S _i)∩A(S _i)=R(S _i)i,<n}

In formula: L _kfor level is divided the each class set drawing;

Step 8: the reachability matrix removal of binary relation of bypassing the immediate leadership to step 7 after level is divided, and remove the binary relation self arriving, and obtain skeleton matrix, and then obtain Multilevel Hierarchical hierarchy Model, consult Fig. 3.

(2) utilize FMECA technology to carry out fault analysis to gained key subsystem and try to achieve critical failure pattern;

Utilize FMECA technology, the fault mode that may exist by the each ingredient of Analysis deterrmination numerically-controlled machine key subsystem and the impact of each fault mode on work of numerical control machine, find Single Point of Faliure, and according to each fault mode severity and each fault mode probability of happening, determine each fault mode harmfulness, for taking correspondingly to remedy innovative approach to eliminate or to alleviate each fault mode harmfulness foundation is provided.Step is as follows:

Step 1: Failure Mode Effective Analysis; Be related in situation at known key subsystem reliability logic, by fault statistics data, the fault mode that key subsystem parts are likely occurred is added up, and analyzes impact and consequence that each fault causes numerically-controlled machine;

Step 2: severity is determined; Final influence degree according to fault mode in fault data to numerically-controlled machine, determines the severity of each fault mode; Consult table 1.

Table 1 system severity is divided

Step 3: fault mode probability of happening is determined; From the fault statistics data of key subsystem, analyze the system failure pattern that obtains.Fault mode probability of happening grade is divided into 5 grades, consults table 2.

The grade classification of table 2 fault mode probability of happening

Step 4: fill in FMECA form; The combined influence that the each fault mode severity of key subsystem and fault mode probability of happening are produced is classified, and by filling in FMECA form, the pattern impact of likely breaking down is carried out to comprehensive comprehensive evaluation; Consult table 3.

Table 3FMECA form

Numbering: analyst's name: design engineer's name: date:

Step 5: harmfulness matrix analysis; Harmfulness matrix horizontal ordinate is fault mode severity grade, ordinate is the probability grade of fault mode, put the distance of initial point by the vertical projection of each fault mode distributed points on diagonal line and carry out comparison harmfulness size, this fault mode harmfulness of the longer explanation of distance is larger.

Specific embodiment

The present invention, take certain serial numerically-controlled machine as example, in conjunction with fault statistics related data, considers fault subsystem correlationship, adopts DEMATEL method, and application digraph and matrix operation obtain fault subsystem combined influence matrix and the degree of correlation; And integrated ISM method, interactional numerous key elements logical relation is carried out to visual representation by Multilevel Hierarchical hierarchy Model, synthesis pertinence and Multilevel Hierarchical hierarchy Model obtain key subsystem.For further determining the critical failure pattern of key subsystem, utilize FMECA technology to carry out HAZAN to key subsystem, thereby find out critical failure pattern, specify reliability improvement emphasis.The system applies of the method is to all important in inhibitings of reliability consideration such as fault diagnosis and location, system reliability improvement.

(1) integrated DEMATEL-ISM method analysis of failure subsystem influences each other, and obtains the fault subsystem degree of correlation and Multilevel Hierarchical hierarchy Model, and then obtains key subsystem;

Step 1: in conjunction with certain serial fault of numerical control machine tool statistical dependence data message, it is carried out to analysis and arrangement and obtain dependent failure statistical form, consult shown in table 4.According to the cause-effect relationship that dependent failure occurs between each fault subsystem, build digraph, consult Fig. 1.

Table 4 numerically-controlled machine subsystem dependent failure statistical form

Step 2: digraph is converted to the direct relational matrix Y that affects

$=\begin{array}{c}1\\ 2\\ 3\\ 4\\ 5\\ 6\\ 7\\ 8\\ 9\end{array}\left(\begin{array}{ccccccccc}0& 6& 4& 1& 3& 0& 2& 1& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 2& 0\\ 0& 5& 0& 0& 0& 0& 0& 1& 0\\ 0& 2& 1& 0& 0& 0& 0& 0& 0\\ 0& 0& 2& 0& 0& 0& 0& 0& 0\end{array}\right)$

Y _ijfor fault subsystem i affects the number of times that affects of fault subsystem j

Step 3: to Y standardization, obtain standardization matrix X

$=\begin{array}{c}1\\ 2\\ 3\\ 4\\ 5\\ 6\\ 7\\ 8\\ 9\end{array}\left(\begin{array}{ccccccccc}0& 0.353& 0.235& 0.059& 0.176& 0& 0.118& 0.059& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0.118& 0& 0& 0& 0& 0.118& 0\\ 0& 0.294& 0.294& 0& 0& 0& 0& 0.059& 0\\ 0& 0.118& 0.059& 0& 0& 0& 0& 0& 0\\ 0& 0& 0.118& 0& 0& 0& 0& 0& 0\end{array}\right)$

Step 4: consider between fault subsystem directly and remote effect relation, can try to achieve combined influence matrix and be

$X=\begin{array}{c}1\\ 2\\ 3\\ 4\\ 5\\ 6\\ 7\\ 8\\ 9\end{array}\left(\begin{array}{ccccccccc}0& 0.395& 0.274& 0.059& 0.176& 0& 0.118& 0.066& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0.014& 0.125& 0& 0& 0& 0& 0.118& 0\\ 0& 0.301& 0.298& 0& 0& 0& 0& 0.059& 0\\ 0& 0.118& 0.059& 0& 0& 0& 0& 0& 0\\ 0& 0& 0.118& 0& 0& 0& 0& 0& 0\end{array}\right)$

Step 5: investigate element T in T _ij, calculate four relevance degrees, consult table 5

Table 5 fourfold correlation degree value

Consult table 5, known by centrad value, electrical system, feed system, tool magazine, digital control system, axis system etc. contact closely with other subsystems.The degree of impact of electrical system is 1.09, and degree of being affected is 0, illustrates that this system only affects other subsystems in dependent failure relation, while there is dependent failure, is the source of trouble with other executive subsystems (parts).All be subject to the direct impact of electrical system in conjunction with the known tool magazine of direct relation matrix, feed system, cooling system, chip removal system, digital control system and axis system, being subject to electrical system to affect larger is tool magazine and feed system.The degree of impact of feed system is 0, degree of being affected is 0.87 to rank first place, and illustrating the most susceptible in all dependent failure subsystems is feed system, and illustrates that this subsystem is affected by other subsystems only, and do not affect other any subsystem impacts, be presentation fault subsystem.Fault rootstock in conjunction with the known feed system of direct relation matrix has electrical system, lubricating system, digital control system, axis system and hydraulic system, wherein digital control system and electrical system on impact larger.The failure rate that is calculated tool holder system by upper analysis meter is that total failare rate cuts its dependent failure rate.Other is as tool magazine, feed system, and cooling system, chip removal system is to be all affected system.

The centrad of each fault subsystem and the reason line of writing music is consulted shown in Fig. 2, the point that centrad and reason degree overlap, be that electrical system, lubricating system and hydraulic system degree of being affected equal 0, for fault " root reason ", and along the symmetrical numerical point of transverse axis, be that tool magazine, feed system, cooling system and chip removal system degree of impact are 0, illustrate that this subsystem must be the presentation fault subsystem being caused by other failure cause.

Draw to draw a conclusion according to above-mentioned analysis:

When electrical system, lubricating system and hydraulic system (claiming temporarily driver sub-system), with other executive subsystems (parts), dependent failure occurs, be the source of trouble;

The failure rate of executive subsystem is as tool magazine, feed system, and cooling system, the failure rate of chip removal system etc. should be removed the dependent failure rate being caused by driver sub-system by total failare rate.

Can determine in dependent failure subsystem and contact subsystem closely with other subsystems by the height of centrad value, can determine the maximum source of trouble by the height of reason degree value.

Step 6: calculate reachability matrix M

By entire system influence matrix

$H=T+I=\begin{array}{c}1\\ 2\\ 3\\ 4\\ 5\\ 6\\ 7\\ 8\\ 9\end{array}\left(\begin{array}{ccccccccc}0& 0.395& 0.274& 0.059& 0.176& 0& 0.118& 0.066& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0.014& 0.125& 0& 0& 0& 0& 0.118& 0\\ 0& 0.301& 0.298& 0& 0& 0& 0& 0.059& 0\\ 0& 0.118& 0.059& 0& 0& 0& 0& 0& 0\\ 0& 0& 0.118& 0& 0& 0& 0& 0& 0\end{array}\right)$

Can determine according to the following formula element value in reachability matrix

wherein λ is given threshold value, for system simplification.For the less system of n value, conventionally without simplification, λ=0 can be set.

Press λ=0 and calculate, can obtain reachability matrix M:

$M=\begin{array}{c}1\\ 2\\ 3\\ 4\\ 5\\ 6\\ 7\\ 8\\ 9\end{array}\left(\begin{array}{ccccccccc}1& 1& 1& 1& 1& 0& 1& 1& 0\\ 0& 1& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 1& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 1& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 1& 0& 0& 0& 0\\ 0& 1& 1& 0& 0& 1& 0& 1& 0\\ 0& 1& 1& 0& 0& 0& 1& 1& 0\\ 0& 1& 1& 0& 0& 0& 0& 1& 0\\ 0& 0& 1& 0& 0& 0& 0& 0& 1\end{array}\right)$

Step 7: above-mentioned reachability matrix is carried out to region division, consult table 6.All belong to the same area by the known all key elements of this table, therefore carry out on this basis level division, consult shown in table 7.

Table 6 reachable set, in advance collection, jointly collect and play initial set table

S i	R(S i)	A(S i)	R(S i)∩A(S i)=A(S i)	B(S i)
					1	1,2,3,4,5,7,8	1	1	1
2	2	1，2，6，7，8	2	?
					3	3	1，3，6，7，8，9	3	?
4	4	1，4	4	?
					5	5	1，5	5	?
6	2,3,6,8	6	6	6
					7	2,3,7,8	1，7	7	?

8	2,3,8	1，6，7，8	8	?
					9	3，9	9	9	9

Table 7 level partition process table

Whole system can be divided into level Four, Π (P ₁)=L ₁, L ₂, L ₃, L ₄={ S ₂,, S ₃, S ₄s ₅, { S ₈, S ₉, { S ₆, S ₇, { S ₁.Now obtain the reachability matrix M' after level is divided

$M^{\&prime;}=\begin{array}{c}2\\ 3\\ 4\\ 5\\ 8\\ 9\\ 6\\ 7\\ 1\end{array}1\left(\begin{array}{ccccccccc}1& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 1& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 1& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 1& 0& 0& 0& 0& 0\\ 1& 1& 0& 0& 1& 0& 0& 0& 0\\ 0& 1& 0& 0& 0& 1& 0& 0& 0\\ 1& 1& 0& 0& 1& 0& 1& 0& 0\\ 1& 1& 1& 0& 1& 0& 0& 1& 0\\ 1& 1& 1& 1& 1& 0& 0& 1& 1\end{array}\right)$

Step 8: the reachability matrix of dividing through level is carried out to backbone and extract, and build the Multilevel Hierarchical hierarchy Model of fault subsystem key element.

Consult Fig. 3, this model is the Multilevel Hierarchical hierarchical model of a level Four, wherein tool magazine (S ₂), feed system (S ₃), cooling system (S ₄), chip removal system (S ₅) be top layer fault subsystem; Axis system (S ₈) and hydraulic system (S ₉) be shallow-layer fault subsystem; Lubricating system (S ₆) and digital control system (S ₇) be middle level fault subsystem; Electrical system (S ₁) be deep layer fault subsystem.Therefore, should guarantee electrical system reliability at the relevant components and parts of Design and manufacture and parts as far as possible.

Find by above-mentioned analysis: the source of trouble that two kinds of methods obtain is consistent with the surface phenomenon of mal-function; Application DEMATEL method can obtain the size of each subsystem degree of impact, degree of being affected, centrad and reason degree, obtain relevancy ranking, determine with other subsystems and contact subsystem and the maximum source of trouble the most closely, and can obtain subsystem fault rate algorithm, but cannot obtain each subsystem fault transport mechanism; And ISM method makes fault delivery relation be able to visualize by setting up Multilevel Hierarchical hierarchy Model, exactly make up the deficiency of DEMATEL method, two kinds of analytical approachs complement one another, and get rid of and have vital role for accurately carrying out fault diagnosis and fault.Because electrical system is deep layer fault subsystem, meanwhile, known in conjunction with centrad and reason degree, electrical system is to contact subsystem and the maximum source of trouble the most closely with other subsystems, can determine that this subsystem is the very important key subsystem in status, need strengthen reliability improvement to its emphasis.

(2) utilize FMECA technology to carry out fault analysis to gained key subsystem and try to achieve critical failure pattern;

This series numerically-controlled machine electrical system mainly comprises various sockets in numerical control tank, relay, contactor, power lead, frequency converter, switching lamp, power supply etc.

Step 1: Failure Mode Effective Analysis.By fault statistics data, the fault mode that key subsystem parts are likely occurred is added up, and analyzes impact and consequence that each fault causes numerically-controlled machine; Consult table 8.

Step 2: severity is determined.Final influence degree according to fault mode in fault data to numerically-controlled machine, determines the severity of each fault mode; Consult table 1.

Step 3: fault mode probability of happening is determined.From the fault statistics data of key subsystem, analyze this subsystem fault pattern that obtains.Fault mode probability of happening grade is divided into 5 grades; Consult table 2.

Step 4: fill in FMECA form.The combined influence that the each fault mode severity of key subsystem and fault mode probability of happening are produced is classified, and by filling in FMECA form, the pattern impact of likely breaking down is carried out to comprehensive comprehensive evaluation.

Based on fault statistics data, this series numerically-controlled machine electrical system is carried out to FMECA analysis, consult table 8.

Table 8 electrical system FMECA table

Step 5: harmfulness matrix analysis.Harmfulness matrix horizontal ordinate is fault mode severity grade, ordinate is the probability grade of fault mode, put the distance of initial point by the vertical projection of each fault mode distributed points on diagonal line and carry out comparison harmfulness size, this fault mode harmfulness of the longer explanation of distance is larger.

According to table 8, can obtain the each fault mode HAZAN of electrical system result, consult table 9.

Table 9 electrical malfunction pattern HAZAN result table

Consult table 9 known, this series numerically-controlled machine electrical system has 12 fault modes, and wherein severity is have 5 of I class, and II class has 3, illustrate that the most fault mode severities of electrical system are very high, can find out that thus electrical system is very large on whole aircraft reliability impact.Consider fault mode probability of happening, the fault mode that harmfulness is larger is followed successively by contactor and damages (D1), relay damaged (G1), power source trip (C2), brake switch damages (A1), stabilized power source damages (C1), and these five fault modes are these series numerically-controlled machine critical failure patterns, should carry out emphatically reliability improvement.These faults mostly are components and parts and damage, and electrical system components and parts are mainly that outsourcing external coordination obtains, and therefore, numerically-controlled machine enterprise should strengthen outsourcing piece purchasing quality management, enter screening of electric components experiment before factory; Design department is in the time selecting components and parts simultaneously, must from preferred handbook catalogue, choose, if must select the components and parts outside catalogue, should be through the investigation of quality department, test, after confirming as reliably, fill into catalogue and just can select, to prevent from causing due to outsourcing piece problem of poor quality the reliability reduction of complete machine.In addition, user also will in use note equipment to strengthen safeguarding, to improve equipment dependability.

Claims (10)

1. a fault of numerical control machine tool analytical approach, is characterized in that, comprises the following steps:

Step 1: integrated DEMATEL-ISM method analysis of failure subsystem influences each other, and obtains the fault subsystem degree of correlation and Multilevel Hierarchical hierarchy Model, and then obtains key subsystem;

In conjunction with fault statistics related data, consider fault correlationship between subsystem, adopt Decision Making Trial and Evaluation laboratory decision experiments chamber analytic approach, application digraph and matrix operation obtain fault subsystem combined influence matrix and the fault subsystem degree of correlation, and apply Interpretative Structural Modeling ISM, interactional numerous key elements logical relation is carried out to visual representation by Multilevel Hierarchical hierarchy Model, the resultant fault subsystem degree of correlation and Multilevel Hierarchical hierarchy Model, obtain numerically-controlled machine key subsystem,

Step 2: utilize FMECA analytical technology to carry out fault analysis to gained key subsystem and try to achieve critical failure pattern;

The fault mode that may exist by the each ingredient of Analysis deterrmination numerically-controlled machine key subsystem, and the impact of each fault mode on work of numerical control machine, find Single Point of Faliure, and according to each fault mode severity and each fault mode probability of happening, determine each fault mode harmfulness, for taking correspondingly to remedy innovative approach to eliminate or to alleviate each fault mode harmfulness foundation is provided.

2. a kind of fault of numerical control machine tool analytical approach according to claim 1, is characterized in that:

Described integrated DEMATEL-ISM method analysis of failure subsystem influences each other, and obtains the fault subsystem degree of correlation and Multilevel Hierarchical hierarchy Model, and then obtains key subsystem, comprises the following steps:

Step 1: according to fault statistics related data, build fault digraph, determine that the fault phase that forms numerically-controlled machine closes set of subsystems S={s _i, i=1,2 ..., n; Wherein S _irepresent that i and other subsystems exist the subsystem of dependent failure, n represents to exist the fault subsystem number of dependent failure;

Step 2: fault digraph is converted into the direct influence matrix Y between all fault subsystems;

Step 3: to the direct influence matrix Y standardization between all fault subsystems;

Step 4: calculate combined influence matrix T

In formula: I is unit matrix, X ^krepresent the k stage remote effect of subsystem i to subsystem j, X is standardization matrix;

Step 5: degree of impact, degree of being affected, centrad, the reason degree of obtaining each fault correlation subsystem;

Step 6: determine reachability matrix M;

Combined influence matrix T only reflects the relation that influences each other and the degree between different faults subsystem, does not consider the impact of fault subsystem on self, therefore needs to calculate the entire effect relation of faults subsystem;

Step 7: Region Decomposition and level are divided: definition subsystem S _ithe set of subsystems of impact is subsystem S _ireachable set R (S _i), affect subsystem S _iset of subsystems be subsystem S _ithe A of collection in advance (S _i), by reachable set R (S _i) and collect in advance A (S _i) calculating, reachability matrix M is carried out to Region Decomposition and level and divides;

Step 8: the reachability matrix removal of binary relation of bypassing the immediate leadership to step 7 after level is divided, and remove the binary relation self arriving, obtain skeleton matrix, and then obtain Multilevel Hierarchical hierarchy Model, thereby obtain key subsystem.

3. a kind of fault of numerical control machine tool analytical approach according to claim 2, is characterised in that:

Described direct influence matrix Y fault digraph being converted between all fault subsystems refers to: failure dependency is taken into account to Y=(y _ij) _{n × n}; Wherein y _ijfor subsystem i affects the number of times that affects of subsystem j, when i=j, y _ii=0, n represents to exist the fault subsystem number of dependent failure.

4. a kind of fault of numerical control machine tool analytical approach according to claim 2, is characterised in that:

The described direct influence matrix Y standardization between all fault subsystems refers to: obtain standardization matrix X

In formula: y _ijfor subsystem i affects the number of times that affects of subsystem j, n represents to exist the fault subsystem number of dependent failure.

5. a kind of fault of numerical control machine tool analytical approach according to claim 2, is characterised in that:

The described degree of impact of obtaining each fault correlation subsystem, degree of being affected, centrad, reason degree refer to: row and the vector of establishing combined influence matrix T are Q, row and the vectorial D of being,

T=[t _ij] _n×n,i,j=1,2,…n

In formula: Q _irepresent degree of impact; D _irepresent degree of being affected; t _ijrepresent the direct and remote effect degree size of subsystem i to subsystem j, t _ij≠ 0, illustrate that subsystem i and subsystem j have the fault relation that influences each other, otherwise irrelevant; N represents to exist the fault subsystem number of dependent failure;

Therefore, Q _i+ D _jcentered by degree, Q _i-D _jfor reason degree:

。

6. a kind of fault of numerical control machine tool analytical approach according to claim 2, is characterised in that:

The entire effect relation of described calculating faults subsystem refers to: calculate entire effect matrix H, its computing formula is:

H=T+I=[h _ij] _n×n

In formula: I is unit matrix, h _ijafter representing to consider that fault subsystem affects self, the direct and remote effect degree size of subsystem i to subsystem j, n represents to exist the fault subsystem number of dependent failure;

By entire effect matrix H, can determine reachability matrix M, order

M=[m _ij] _n×n,i,j=1,2,…n

M _ijvalue according to the following formula:

wherein λ is given threshold value, and the size of λ directly affects reachability matrix formation and follow-up hierarchical structure is divided; For the less system of n value, without simplification, λ=0 is set;

M _ijbe illustrated under given threshold value λ, whether subsystem i there is impact to subsystem j, if h _ij> λ, represents to exist impact, m _ijvalue is 1; If h _ij≤ λ, represents not exist impact, m _ijbe 0.

7. a kind of fault of numerical control machine tool analytical approach according to claim 2, is characterised in that:

Described Region Decomposition refers to: by R (S _i) ∩ A (S _i) be defined as common collection, common collection R (S _i) ∩ A (S _i)=A (S _i) key element be defined as initial set B (S), for the S in B (S) _iand S _jif, R (S _i) ∩ A (S _j)=Φ, S _iand S _jdo not belong to the same area, otherwise be the same area, decomposition that so can feasible region; The result of Region Decomposition can be designated as P (S)=P ₁, P ₂... P _k, wherein P _kit is the set of subsystems in k relatively independent region;

Described level is divided and is referred to: for the same area P ₁, obtain and meet R (S successively _i) ∩ A (S _i)=R (S _i) subsystem, find out each class set, be expressed as

L ₁={S _i/S _i∈P ₁-L ₀,R(S _i)∩A(S _i)=R(S _i)i,=1,…2,n}

L ₂={S _i/S _i∈P ₁-L ₁,R(S _i)∩A(S _i)=R(S _i)i,<n}

…

L _k={S _i/S _i∈P ₁-L ₁-…-L _k-1,R(S _i)∩A(S _i)=R(S _i)i,<n}

In formula: L _kfor level is divided the each class set drawing.

8. a kind of fault of numerical control machine tool analytical approach according to claim 1, is characterised in that:

Described utilize FMECA analytical technology to carry out fault analysis to gained key subsystem to try to achieve critical failure pattern, comprise the following steps:

Step 1: Failure Mode Effective Analysis; Be related in situation at known key subsystem reliability logic, by fault statistics data, the fault mode that key subsystem parts are likely occurred is added up, and analyzes impact and consequence that each fault causes numerically-controlled machine;

Step 2: severity is determined; Final influence degree according to fault mode in fault data to numerically-controlled machine, determines the severity of each fault mode;

Step 3: fault mode probability of happening is determined; From the fault statistics data of key subsystem, analyze and obtain this subsystem fault pattern probability;

Step 4: fill in FMECA form; The combined influence that the each fault mode severity of key subsystem and fault mode probability of happening are produced is classified, and by filling in FMECA form, the pattern impact of likely breaking down is carried out to comprehensive comprehensive evaluation;

Step 5: harmfulness matrix analysis; Harmfulness matrix horizontal ordinate is fault mode severity grade, ordinate is the probability grade of fault mode, put the distance of initial point by the vertical projection of each fault mode distributed points on diagonal line and carry out comparison harmfulness size, this fault mode harmfulness of the longer explanation of distance is larger.

9. a kind of fault of numerical control machine tool analytical approach according to claim 8, is characterised in that:

The severity of described definite each fault mode refers to: severity is divided into 4 ranks, be respectively I critical fault, II catastrophic failure, III generic failure, IV minor failure, wherein I grade failure effect damages for causing numerically-controlled machine, II grade failure effect is to cause heavy economic losses or cause the numerically-controlled machine of mission failure badly damaged, III grade failure effect is to cause certain economic loss or cause the numerically-controlled machine of task delay or degradation slightly to damage, IV grade failure effect is for being not enough to cause above-mentioned 3 kinds of consequences, but can cause unscheduled maintenance or repairing.

10. a kind of fault of numerical control machine tool analytical approach according to claim 8, is characterised in that:

Described analysis obtains this subsystem fault pattern and refers to: fault mode probability of happening grade is divided into A, B, C, D, an E5 grade, represents respectively often to occur, sometimes occur, occur once in a while, seldom occur, seldom occur; Fault mode probability of happening feature be followed successively by fault mode probability of happening account for more than 20% high probability of total failare probability, be greater than total failare probability 10% or be less than or equal to its middle equiprobability of 20%, be greater than 1% of total failare probability and be less than or equal to its 0.1% of total failare probability that seldom occurs, is greater than of 10% and be less than or equal to its unlikely generation of 1%, account for total failare probability 0.1% or following almost nil.

Images